[HN Gopher] Modular Cognition
___________________________________________________________________
Modular Cognition
Author : the-mitr
Score : 14 points
Date : 2022-03-20 08:04 UTC (1 days ago)
(HTM) web link (aeon.co)
(TXT) w3m dump (aeon.co)
| smegsicle wrote:
| pretty fun talk by the same guy:
|
| What Bodies Think About: Bioelectric Computation Outside the
| Nervous System - NeurIPS 2018
|
| https://yewtu.be/watch?v=RjD1aLm4Thg
| mateo1 wrote:
| That's absolutely incredible, thank you for posting it.
| jorgeleo wrote:
| Not really following this article:
|
| "Likewise, if two mouse embryos are mushed together like a
| snowball, a single, normal mouse results"
|
| And
|
| "This is intelligence in action: the ability to reach a
| particular goal or solve a problem by undertaking new steps in
| the face of changing circumstances"
|
| I write software that can account for changes in context and
| still reach a particular goal. This is not new, creating TCP from
| UDP where the network can constantly be changing is not something
| that I will call "Intelligent". The guy(s) that created the
| original algorithm are, but the network packets themselves are
| not.
|
| There are differences between intelligence and just following
| well crafted procedures.
| jotm wrote:
| That _is_ (very limited) intelligence imo. It 's nothing more
| than a very complex, ever increasing set of procedures plus a
| huge amount of data and a fast processor.
|
| I mean, the more you know about something, the less "magic" it
| seems. You know algorithms, the author may not. Pretty sure
| someone from the 1700's would think a smartphone with some
| personal assistant software on it was intelligent (even if
| possessed), for example.
___________________________________________________________________
(page generated 2022-03-21 23:02 UTC)